Why is computer-based go interesting?
Go is the last remaining significant board game to await mastery by the computer. In chess, the most complex Western board game, the number of humans stronger than the computer is now precisely zero: a computer defeated the reigning world champion in the historic 1997 match documented here. By contrast, in go, the number of humans stronger than the computer is perhaps ten million.
But why are computers so weak at go?
The answer lies equally in the greater complexity of go, as well as the fact that humankind has not organized itself well in attacking the computer go challenge. The brightest computer scientists of the day and the largest computer companies with access to the most advanced resources joined forces to solve the computer chess problem. Computer go, on the other hand, is still largely the domain of the lone programmer/go player working in his study.
But....so what?
Creating a strong go program is clearly a challenge worthy of our best efforts, in and of itself. But in addition, successfully solving this problem is certain to teach us a great deal about new approaches to problem-solving with the computer. Also, the existence of a strong go-playing program will be of immense help in go education, thus aiding in the popularization of this fascinating game.
|
|
|
Copyright (c) 2000, Intelligent Go Foundation. All rights reserved. |